I suppose by now the 2010 Census is passe, or even denounced. But I’m bit behind the times, so I’m still thinking about it, and it’s taken me this long to parse some of the results. By which I mean I’ve made a list of the downtown populations of MSAs of 1 million or more as reported in the 2000 and 2010 censuses, along with the metro and central city populations.
Specifically, I was a bit skeptical of some claims that downtowns have been resurgent in the last decade. The condo craze was an undisputed fact, mostly a result of cheap and easy credit, that mostly affected the central areas of cities. But was it big enough to really have an effect on population? And was it nationwide or limited to major cities like San Francisco or trendy cities like Portland?
The first two problems with answering those questions are both of definition. First, the definition of the word downtown has shifted subtly over time, so that now it seems to have different meanings depending on the type of place it is in. The word originally referred to the older section of the city, which was located down hill from the newer suburbs. This was quickly supplanted with a meaning of referring to the main business district of the city, which is the definition commonly found in dictionaries. Now, however, it is often used to refer to a relatively dense or mixed-use area, or an area with a New Urbanist form, which could be pretty much the same as any strip mall, only with parking in back.
So what counts as downtown? In St Paul, the Upper Landing area was developed with a high-density urban form in an area close enough to the traditional downtown that it could be considered a new urban appendage. But traditionally this area would not have been considered downtown, at least not since Little Italy was cleared. So should its growth be credited to downtown or to the rest of the city?
I took the coward’s route – I tried to use local definitions of downtown. First, if I could find a media report – blog or newspaper – that defined a city’s downtown, I let them do the work for me. If I couldn’t find a media report, I used the municipality’s definition – most cities have an area within boundaries that they’ve called Downtown, even if just for economic development. There were a handful of cities where I just trusted Wikipedia, because I couldn’t find a municipal definition. Also some of the bigger cities I made up definitions of Greater Downtown Areas because the official definition was too small an area. For example Chicago’s Loop is defined as south and east of the Chicago River, west of the lake and north of Roosevelt, making a bit more than half the land area of Minneapolis’ official inside-the-freeways, west of the river Downtown (1.49 sq mi for the former and 2.71 sq mi for the latter).
Once I had a definition of a downtown, I used the New York Times 2010 Census map to get the 2010 population and rate of growth, from which I extrapolated the 2000 population. Because official downtown definitions don’t always follow the boundaries of census tracts, I just tried to approximate as best I could. Feel free to download my data for definitions of particular cities and the actual census tracts I used.
Now that I’ve provided a convoluted introduction, let’s get to the results. The answer is yes, downtowns nationwide grew in the last decade. Only 6 of the metros with a million-plus population (hereafter referred to as MPMs, for Million-Plus Metros) had downtowns that didn’t grow in the last decade, out of 51 total. Just when you thought we were going to get to some graphs, that brings up another peculiarity. As Twin Citians know, some regions have more than one regionally important downtown. In those cases I combined the populations of both downtowns to gauge the regional downtown population. That of course involved calculating them separately, whereupon I found that in Riverside-San Bernardino, one of the downtowns grew while the other shrank (albeit by an insignificant 30 people). So the combined downtown population grew by 5%, but San Bernardino grew by 9% while Riverside shrank by -.27%. Both of them appear to actually be suburbs of LA anyway, so maybe it doesn’t matter, but it’s something to keep in mind when the graphs start flying.
Let’s kick it off with a straight up graph of population change, absolute and percent, in the 51 MPMs:
This graph is pretty hard to read because of some mega outliers, specifically Manhattan, which grew by 49,174 (the next greatest growth was LA 15,445), and Dallas-Fort Worth, which grew by 306% (the next greatest percent growth was Charlotte with 134%).
The average growth for all MPMs was 41%, which sounds impressive until you hear the average absolute growth, which was only 4,493 people. That’s chicken scratch next to the average growth for the entire metro area across the 51 MPMs, which was 315,214. So let’s look at that. Here’s a graph of how downtown growth rates compared with metro area growth rates and central city growth rates:
I cut off the Dallas-Fort Worth downtown growth rate for this one so the other rates would show up better. As you can see, downtowns had much stronger growth rates overall than metro areas and especially central cities. However, that pattern isn’t across the board – around 10 metros on the left side of the chart had greater metro area and central city growth rates than downtown growth rates. I classified the metros by census bureau region, and if you look at the average across regions, you start to see some different patterns:
On this one, the weird result is the Northeast’s Central City, which shows a percent decrease but an absolute increase. It’s all New York’s fault – because it increased by 166,855 people (the largest central city increase), it drives the average way up. But for a city the size of New York, that’s only a 2% increase in population, not enough to offset the large percent declines of Buffalo, Pittsburgh and Rochester (-11%, -9%, and -4% respectively). I guess that’s my warning not to read the percent columns as an expression of the numerical columns.
Anyway, the regional differences become clear. The South had about twice the average rate of downtown population change of any other region, but it also had a much lower numerical change. This likely accounts for the impressive-sounding rate of change for all cities, of which the South makes up close to half. Still, it’s impressive that a region that had almost nothing that could be considered urban now has its first taste of city life. Conversely, the Northeast had the lowest average downtown population rate of change and the highest average numerical change, reflecting the higher existing populations in the downtowns of that region.
Looking at the average downtown population across regions seems to bear this theory out, as the South has the lowest average population in 2000:
And again the South comes out poorly when looking at the cities with the 10 largest downtown populations, with only Raleigh representing the vast region:
But Raleigh also exposes the methodological weakness of this whole venture, as does the top 10 table in general. The only definition I could find of Raleigh’s downtown runs east to Raleigh Blvd, so it includes a couple square miles of streetcar suburb. If I were to define Downtown Raleigh, the eastern border wouldn’t run past East St, so the definition I used is about twice the actual size even in my own estimation.
On the other side, you may have noticed that I used the entire borough of Manhattan for downtown New Y0rk. You may disagree with that decision, and if so you’re not alone – I do, too. The problem is that the part of Manhattan sometimes called “Downtown” actually has fewer jobs and office space than Midtown Manhattan, so each area would conflict with some definition. Since even the northern tip of the island has characteristics that could be considered a downtown in any other American city, I decided to look at the borough as a whole. That approach is probably unfair to other New York business districts, such as Downtown Brooklyn, that probably could have qualified for this study. It also makes for a freaking huge outlier, although Lower and Midtown Manhattan are also outliers, with six times the 2010 population of the next biggest city, and 7,000-15,000 greater increase in population than the next largest increase (download my data if you’re curious about specifics).
Manhattan was the only city where I included a greater downtown area for comparison’s sake, although as mentioned above I did create them for Chicago, Miami and San Francisco – that’s the kind of inconsistency that comes from dealing with this much data in your spare time. My policy of generously allowing cities to define their own downtowns created its own problems, especially when looking at the percent of metro area population that lives downtown, where four Southern cities appear in the top ten:
Note that New York has been omitted from the two above charts, and that while in general downtowns’ share of metro area population went up, it didn’t go up as much as the two charts seem to show due to my neglect to maintain the same scale for the right axis.
So after a decade that was widely hyped for downtown growth, only seven cities have downtowns that contain 1% or more of their metro’s populations. While the ubiquity of downtown growth may have been exceptional, the growth itself may not have been significant, or significant enough to make a difference regionally. Downtown growth mostly beat regional growth in the relatively sluggish decade (see the chart Downtown, Central City, and Metro Area Growth Rates above), and downtowns as generally small portions of metro areas maybe shouldn’t be expected to ever contain a significant portion of the metro population. But central cities grew much less consistently (as we know here in the Twin Cities), with 21 central cities shrinking, and of the 30 that grew, only maybe a dozen did so without annexation or greenfield development. Moreover, as I attempt to show on the most confusing chart I’ve ever made, the vast majority of central cities lost share of metro area population, that is, the suburbs grew faster than they did:
So while downtown growth was a national trend, we’re not talking about an overthrow of the suburban realm quite yet. Still, even a small shoot of urban growth is encouraging in the vast mire of suburbanism that has festered for the past few decades. Things change slowly in our litigation-fueled, Nimbyful society, and there are indications that demand for urban places exceed the supply. The fringe still seems to be freckled with vacancy while apartments rise in the central city, which indicates a lot taller bars in the charts of downtown growth after the 2020 census.
If you care to dig through my data, here is the xlsx. Feel free to dispute the downtown definitions – more likely than not I’ll agree with you.